Philosophiae Doctor (PhD) Thesis 2018:105
Elia Ciani
Neuroendocrine regulation of pituitary gonadotropins during puberty in Atlantic salmon parr with focus on melatonin and Gnrh systems
Nevroendokrin regulering av gonadotropiner i atlantisk laks parr med fokus på melatonin og Gnrh systemer
Norwegian University of Life Sciences Veterinary Medicine
Neuroendocrine regulation of pituitary gonadotropins during puberty in Atlantic salmon parr with focus on melatonin and Gnrh systems
Philosophiae Doctor (PhD) Thesis Elia Ciani
Department of Basic Sciences and Aquatic Medicine Faculty of Veterinary Medicine
Norwegian University of Life Sciences Oslo 2018
2
Thesis number 2018:105 ISSN 1894-6402 ISBN 978-82-575-1770-0
3
Acknowledgements
During these three years of PhD at the Norwegian University of Life Sciences, I had the luck of being guided from wise supervisors, surrounded by talented researchers, and supported by good friends. I would like to start by acknowledging my PhD committee for evaluating this thesis and to express my gratitude to my supervisor Finn-Arne Weltzien for always believing in me and for granting me the opportunity to be part of his lab. Prof, I highly appreciate your guidance during these years and your moral support during the last days of thesis writing. I want to thank my co-supervisors Berta Levavi-Sivan, for making me feel at home during my permanence in Israel and for the productive collaboration with her group. To my co-supervisor Romain Fontaine, thank you for all that you taught me in the lab, for your scientific guidance and for always finding some time to discuss with me, despite you’re one of the busiest persons I’ve ever met. I would like to acknowledge the head of department, Ole Taugbøl for all his support.
All the people from the Weltzien lab deserves my deepest gratitude: Eirill, thank you for all your dedication in making the IMPRESS project such a stimulating program of scientific research and personal growth. Gersende, your skills and constructive collaboration have been much appreciated in the production of the melatonin paper.
Khadeeja thanks for your positive spirit, it is a pleasure to have you as a colleague. To Kjetil, thank you for your help during these years, especially for the non-stop sampling at Ims. Christiaan thank you for teaching me that tulips are much more complex and interesting than what I was expecting. Kristine, simply the best. I much appreciated your hundreds of histological analyses, sharp comments, revisions, and advices. They really helped me to improve the quality of my papers. Lourdes thank you for your cheerful presence and all the help in the lab. I would never have been able to finish in time my analysis without the help of Nouri with qPCR. Thank you also for your contagious laugh, much appreciated during lab work. Susann I’ve been really happy to share this PhD experience with you.
Thanks to all my colleagues at NMBU for making such a great work environment.
Special thanks to Ian Mayer for all the help with my experiments and for the useful scientific discussions.
I would like to thank all the members of Levavi-Sivan’s lab: Krist, Lian, Michal, Yaron and, in particular, Naama for teaching me the secrets behind receptor
4
pharmacology and cell culture, and for the delicious dinners. Yaron, thank you for helping me to actually reach the place I want to go when using Israelian public transportation.
As part of IMPRESS, I had the chance to be part of a group of incredibly talented ESR’s and, indeed, the most anticipated periods of the last three years have been the IMPRESS meetings. Thank you Aurora, Ben, Carina, Carlos, Chris, Daan, Elena, German, Hannah, Ishwar, Maria, Mitch, Rasheed, Sahana, Sophia, those years with you have been really amazing and for sure, we have the best “last supper” picture ever.
Daan, it was really fun to have you as lab neighbour in Oslo, and I much appreciated your help during sampling and compiling boring burocratic papers.
Living in a foreign country always gives you a bit of home sickness. I feel really grateful to my “Italian ring” + guests: Ale thank you for all the nights out, the travels and the beers together. To the most efficient organizer of the group, Chiara thank you for taking care of all our trips together. I’m glad that Oslo gave me the chance to meet my new sister, Floriana. Fra I missed you a lot since you left for Italy and I really had a tough time to remember where the Illegal Burger is, ever since. Gian Enrico, you prepared the best frokost ever. Giulia you’ve been a good flatmate, colleague and friend, I would like to thank you for hosting me on your (and Chiara’s) couch in my visits back to Oslo. To the most assiduous training partner, Leo, thanks for the bike. Iris, thanks for the nice discussions and wine glasses.
To all my friends around the world, in Italy and to the gang of Abruzzo: Davide X, Domenico, Fratta & Lucia, Germano Mosconi, the Cadderis, Marco & Gloria, Max
& Sabina, Pavone, Piero, Robberta, Samuele, Sara & Pessa, Silvia, now that I am constantly flying around, I appreciate even more the time we spend together. In particular I want to acknowledge my siblings Daniele and Sara. Even if we don’t meet often now, you’ve always been there to keep my mental health intact, or to laugh at me, I appreciate both. Despite years have passed by, it’s always a pleasure to be in contact with the Lurie luride, the best group of marine biologists and fattoni ever. Special thanks to Giorgia Di Muzio for her beautiful drawings of Salmo salar parr.
Since most of my family cannot read English, I hope my English speaking readers wouldn’t mind for the next few lines in Italian to acknowledge my family. Mi siete stati sempre vicino e mi date la forza di andare Avanti. Grazie a tutta la mia famiglia, a tutte le persone che negli anni sono diventate parte di essa, e a chi non c’è piu ma è sempre
5
nei miei pensieri: Angela, Ada, Nonno Vincenzo e Zia Rina; Nonno hai visto, adesso sono finalmente Dr. Ciani. Anna, Ilenia, Paolo e Valentina; Pà grazie per avermi campato. Alberto, Carla, Francesco, Giovanni, Maria, Marina, Paola, Sergio e Vittorio; Sergiuz, sono sempre molto felice di averti conosciuto, anche se combini i casini. Cinzia, Nonna Vincenzina, Zia Gina e Zio Do; Zio, lo so che tu continui a guidarmi da su. Alessia, La Nonna Rosi e Manola; Nonna, sei la donna più forte del mondo.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No [642893] and the Norwegian University of Life Sciences.
And the earth becomes my throne I adapt to the unknown Under wandering starts I’ve grown By myself but not alone Anywhere I roam, where I lay my head is home (Wherever I may roam – Metallica)
Salmo salar parr.
6
Contents
Acknowledgements ... 3
List of papers ... 8
Abbreviations ... 9
Species nomenclature ... 11
Norsk sammendrag ... 13
Introduction ... 15
General background ... 15
Model species: Atlantic salmon ... 16
The brain-pituitary-gonad (BPG) axis ... 19
Pituitary morphology ... 20
GnRH ... 22
GnRH receptors... 25
Intracellular signaling pathways activated by Gnrh ... 27
Gonadotropins... 29
Melatonin ... 30
Melatonin receptors ... 34
Testes morphology ... 35
Spermatogenesis ... 36
Aims of the study ... 38
Methodological considerations ... 39
qPCR ... 39
Fluorescent in situ hybridization ... 42
Reporter gene assay ... 44
Results ... 48
Paper I ... 48
Paper II ... 48
7
Paper III ... 48
Discussion ... 50
General conclusion ... 52
Future Perspectives ... 53
References ... 54
Appendix (my papers) ... 82
8
List of papers
Paper I
Sexual maturation in Atlantic salmon male parr is triggered both in early spring and late summer under standard farming condition.
Elia Ciani, Kristine von Krogh, Rasoul Nourizadeh-Lillabadi, Ian Mayer, Romain Fontaine, Finn-Arne Weltzien.
Manuscript
Paper II
Expression of Gnrh receptor gnrhr2b1 exclusively in lhb-expressing cells in Atlantic salmon male parr
Elia Ciani, Romain Fontaine, Rasoul Nourizadeh-Lillabadi, Eva Andersson, Jan Bogerd, Kristine von Krogh, Finn-Arne Weltzien
Manuscript
Paper III
Melatonin receptors in Atlantic salmon stimulate cAMP levels and show season- dependent circadian variations in pituitary expression levels
Elia Ciani, Romain Fontaine, Gersende Maugars, Naama Mizrahi, Ian Mayer, Berta Levavi-Sivan, Finn-Arne Weltzien
Manuscript submitted to Journal of Pineal Research
9
Abbreviations
This thesis applies the nomenclature recommendations from MGI (http://www.informatics.jax.org/mgihome/nomen/gene.shtml) and ZFIN (https://wiki.zfin.org/display/general/ZFIN+Zebrafish+Nomenclature+Conventions), where mammalian genes are italicized, first letter uppercase (e.g. Fshb); fish genes are italicized, all letters lowercase (fshb); mammalian (or unspecified) proteins are non- italicized, all letters uppercase (FSH); and fish proteins are non-italicized, first letter uppercase (Fsh).
11KT 11-keto testosterone
5-HT 5- hydroxytryptamine (serotonin) AANAT arylalkylamine N-acetyltransferase AC adenylate cyclase
ACTH adrenocorticotropic hormone
BK- KCa big conductance calcium-activated potassium channels cAMP cyclic adenosine monophosphate
D darkness
DAG diacylglycerol DD constant darkness
DHP 17α,20β-dihydroxy-4-pregnen-3-one
E2 17β-estradiol
ER endoplasmatic reticulum
ERK extracellular signal related kinases FSHb follicle-stimulating hormone β subunit FSHR follicle-stimulating hormone receptor GABA gamma-aminobutyric acid
GAP GnRH-associated peptide
GH growth hormone
GnIH gonadotropin inhibitory hormone GnRH gonadotropin releasing hormone
GnRHR gonadotropin releasing hormone receptor GPa gonadotropin subunit α
GtH gonadotropin hormone HCG human chorionic gonadotropin HIOMT hydroxyindole-O-methyltransferase BPG brain-pituitary-gonad axis
IGF-I insulin-like growth factor
IMPRESS improved production strategies for endangered freshwater species IP3 inositol triphosphate
JNK jun n-terminal kinases
KCa calcium-activated potassium channels
L light
LHb luteinizing hormone β subunit LHR luteinizing hormone receptor LL constant light
10 MAPK mitogen activated protein kinase MSH melanocyte stimulating hormone Mtnr melatonin receptors
NPY neuropeptide Y PD pars distalis PI pars intermedia
PIP3 phosphatidylinositol (3,4,5)-trisphosphate PKA protein kinase A
PKC protein kinase C PLC phospholipase C PN pars nervosa
PPD proximal pars distalis PRL prolactin
RPD rostral pars distalis
T testosterone
TSH thyroid stimulating hormone UTR untranslated region
VSCC voltage sensitive calcium channels
11
Species nomenclature
Throughout this thesis, the common name of a species is followed by the scientific name upon its first use. Only the common name will be given in consecutive use, and in case no common name exist in English, only the scientific name will be used African catfish Clarias gariepinus
Atlantic cod Gadus morhua
Atlantic croaker Micropogonias undulates
Ayu Plecoglossus altivelis
Bluefin tuna Thunnus thynnus
Chum salmon Oncorhynchus keta
Coho salmon Oncorhynchus kisutch European eel Anguilla anguilla European sea bass Dicentrarchus labrax
Goldfish Carassius auratus
Haddock Melanogrammus aeglefinus
Japanese eel Anguilla japonica Lake whitefish Coregonus clupeaformis Masu salmon Oncorhynchus masou
Medaka Oryzias latipes
Nile tilapia Oreochromis niloticus Pacific herring Clupea harengus pallasi
Pike Esox lucius
Rainbow trout Oncorhynchus mykiss
Sea bream Sparus aurata
Senegalese sole Solea senegalensis Sockeye salmon Oncorhynchus nerka Spiny dogfish Squalus acanthias Spotted green pufferfish Tetraodon nigroviridis White sucker Catostomus commersonii
Zebrafish Danio rerio
12
Summary
The pituitary gland, gonadotropins, together with neurohormones like gonadotropin-releasing hormone (Gnrh) and melatonin are key players in the control of vertebrate reproduction. In teleosts, despite the numerous studies conducted on the role of these hormones, relatively little is known about the functions of the multiple paralogs of their specific receptors. This study used male parr as a model for investigating the neuroendocrine control of sexual maturation in Atlantic salmon.
First, to identify the onset of maturation and discriminate between maturing and non-maturing fish, testes histology was analysed in combination with gonadosomatic index, gonadotropin expression and plasma steroid levels. The results indicated that sexual maturation can be triggered as early as six months after hatching, in early autumn. Six Gnrh receptor paralogs where found expressed in male parr pituitaries, but only one of them, gnrhr2b1, showed a peak in maturing fish in concomitance with higher gonadotropin expression levels. In situ hybridization localized gnrhr2b1 mRNA specifically to lhb-producing cells, suggesting this receptor to be involved in the regulation of lhb-expression. The third part of this study focused on the characterization of melatonin receptors, considering the importance of melatonin as the “time keeping”
molecule in vertebrates. In silico studies identified five genes encoding putative functional melatonin receptors, phylogenetically clustered in three subtypes, 1A, 1Al and 1B. Pharmacological characterization of cloned receptors, Mtnr1Aaα, Mtnr1Ab and Mtnr1B, proved their functionality in vitro. Three genes, mtnr1aaβ, mtnr1ab and mtnr1b, were expressed in the pituitary, showing intense daily fluctuations in spring, but not in autumn, indicative of important seasonal differences.
The results presented in this thesis add to our understanding on gonadotropin control from Gnrh, suggesting that, in Atlantic salmon, differential regulation may occur through specific receptor paralogs expressed in different cell types. The detection of melatonin receptors in the pituitary gland, showing variation in expression depending on season, strongly advocates for a direct involvement of melatonin in one (or more) of the seasonal processes regulated from the pituitary gland.
13
Norsk sammendrag
Gonadotropiner fra hypofysen, folikkelstimulerende hormon (Fsh) og luteiniserende hormon (Lh) regulerer kjønnsmodning hos alle vertebrater. Utskillelse av Fsh og Lh styres hovedsakelig av nevrohormoner fra hypotalamus, med gonadotropinfrisettende hormon (Gnrh) som det viktigste. Et annet nevrohormon, melatonin, har også en sentral rolle men eksakt hvordan det påvirker kjønnsmodningen er ikke kjent.
I PhD arbeidet har jeg benyttet modnende hannfisk (parr) av atlantisk laks for å studere den nevroendokrine kontrollen av kjønnsmodning. For å skille mellom modnende og ikke-modnende fisk ble det foretatt histologiske analyser av testiklene, sett i sammenheng med gonadosomatisk indeks, samt genuttrykk for Fsh og Lh og plasmanivåer av steroidhormoner. Resultatene viste at kjønnsmodningen kan starte allerede seks måneder etter klekking, tidlig på høsten. Det ble videre identifisert seks paraloge gener av Gnrh-reseptorer i hypofysen hos parr, men bare den ene reseptoren, gnrhr2b1, viste en signifikant økning hos modnende fisk. Dette resultatet samsvarer også med den observerte økningen i genuttrykket til de to gonadotropinene. In situ hybridisering lokaliserte gnrhr2b1 mRNA spesifikt til lhb-produserende celler, noe som kan tyde på at denne reseptoren er involvert i reguleringen av lhb-ekspresjon.
Videre karakteriserte jeg laksens melatoninreseptorer, og betydningen av melatonin som en "time-keeper" i forbindelse med kjønnsmodningen. In silico studier identifiserte fem gener kodende for potensielle melatoninreseptorer. Disse reseptorene ble fylogenetisk gruppert i tre undergrupper, 1A, 1Al og 1B. Farmakologisk karakterisering av de klonede reseptorene, Mtnr1Aaα, Mtnr1Ab og Mtnr1B, viste at disse er fullt funksjonelle in vitro. Tre gener, mtnr1aaβ, mtnr1ab og mtnr1b, var uttrykt i hypofysen, og om våren viste disse reseptorene daglige svingninger i genuttrykk. Om høsten derimot, ble det ikke observert slike svingninger i genuttrykket, noe som indikerer viktige sesongavhengige forskjeller.
Resultatene som presenteres i denne avhandlingen er med på å øke vår forståelse av hvordan gonadotropinene påvirkes av Gnrh i atlantisk laks og indikerer at den ulike reguleringen av gonadotropinene kan skyldes at spesifikke reseptorer er utrykt i ulike celletyper. Identifiseringen av melatoninreseptorer i hypofysen og deres sesongavhengige uttrykksmønstre peker i retning av en mulig rolle for melatonin i
14
direkte påvirkning av en eller flere av de sesongavhengige prosessene kontrollert av hypofysen.
15
Introduction
General background
The Atlantic salmon (Salmo salar) is an iconic anadromous species, with a complex life cycle divided between rivers and ocean (reviewed in Klemetsen et al., 2003). The bond between salmon and people in Norway is centuries old, as part of the economy and culture of the nation (Fig. 1). In the last decades however, the survival of wild salmon has been threatened from several factors linked in a direct or indirect way to human activities including habitat loss or alterations, overfishing and genetic pollution (Jonsson and Jonsson, 2004; Kennedy and Crozier, 2010; Lundqvist et al., 2008; McGinnity et al., 2003). Despite efforts aimed at improving the quality of wild stocks, the number of individuals returning to the cost of Norway every year has decreased 55% since the 1980s. Norwegian wild salmon populations are estimated to 470.000 individuals, compared to the 1.180.000 tonnes of farmed salmon produced only in 2016 (Thorstad and Forseth, 2017). The decline in wild populations is not restricted to Norwegian stocks, and significant decline has been reported also in Spanish and French stocks (Horreo et al., 2011; Le Cam et al., 2015).
Fig. 1 Figgjo rune stein. This stone, exposed at the archaeological museum of Stavanger (south west Norway) and dated to 1100 AD, carries runic inscription about fishing rights in the “salmon river” Figgjo, testifying the antiquity and the importance of the relationship between people and salmon in Norway. The broodstock used in this thesis derives from wild caught salmon from the same river, Figgjo (photo E. Ciani).
My PhD project is part of an ambitious european project named IMPRESS (Improved production strategies for endangered freshwater species; www.impress- itn.eu), which aims at improving restocking techniques for three threatened European diadromous species: Atlantic salmon, European eel (Anguilla anguilla) and Sturgeons (fam. Acipenseridae). A deeper knowledge on regulatory mechanisms behind hormonal
16
and environmental control of reproduction, governed through the brain-pituitary- gonadal (BPG) axis and the circadian system, is necessary to improve the quality of restocking procedures and conservation strategies.
Increased activity of the BPG axis is a key element for the acquisition of reproductive competence and the transition between juveniles and adults in vertebrates, including teleosts fish (Weltzien et al., 2004). The release of gonadotropin- releasing-hormone (Gnrh), stimulate the production of two gonadotropins, luteinizing hormone (Lh) and follicle-stimulating hormone (Fsh) from the pituitary gland, which in turn regulate steroidogenesis and gametogenesis at the level of the gonads (for review see Levavi-Sivan et al., 2010; Schulz et al., 2010; Zohar et al., 2010). Melatonin, the “time keeping” molecule produced from the pineal gland, is involved in the regulation of many seasonal physiological and behavioural processes, including reproduction, acting at different levels of the BPG axis (for review see Falcón et al., 2010). In teleosts however, the details behind the mechanism regulating the differential expression, synthesis and release of Lh and Fsh, as well as the direct effects and cellular target of melatonin in the pituitary, are still not well understood.
Model species: Atlantic salmon
Teleost fish dominate both marine and freshwater environments, represented by almost 30 000 species and accounting for nearly half of all extant vertebrate species (Wootton, 1990). The subfamily Salmoninae comprises about 30 species in seven genera including the genera Salmo, Oncorhynchus and Salvelinus (Klemetsen et al., 2003). Belonging to the genus Salmo, the Atlantic salmon (S. salar) is characterized by a significant plasticity in its life cycle and reproductive strategies (reviewed in Aas et al., 2011; Behnke et al., 2010; Fleming, 1996; Thorpe et al., 1998).
Atlantic salmon are anadromous fish, migrating from oceans to reach spawning sites in rivers in the autumn. The eggs hatch in the following spring and the newly hatched fish (alevins), stay in the gravel for the first 3 to 8 weeks using their yolk sack for nutrition before emerging (as fry) to initiate their first feeding. The moment of emersion from the gravel has to be synchronized with food availability, this is therefore one of the environmental parameters shaping the time of spawning in each river (Heggberget, 1988). The freshwater adapted juveniles (parr, Fig. 2), remain in rivers for a period ranging from 1 to 8 years, depending on both genetic factors, latitude, and
17
environmental conditions. Prior migrating to sea, the fish undergo a process of morphological, physiological, and behavioural transformation, named smoltification, preparing it for the life in sea water. At this stage, the smolts may weight from 10 to 80 g and measure from 7 to 30 cm in total length. After entering to sea, those fish face a second critical “bottleneck” for survival, coping with a new environment, different predators and new types of prey (Klemetsen et al., 2003). The smolts spend 1 to 5 years at sea in foraging areas over a wide area in North Atlantic Ocean before returning to home rivers for spawning. During this migration, the fish display high variability in body weight within and among population, ranging from 1 to 25 kg of weight and 45 to 135 cm of total length.
Fig. 2 Atlantic salmon parr. Identifiable from the “parr marks”, dark camouflaging striped along their sides.
The fish in the centre is showing early signs of smoltification having lost the parr marks and displaying a coloration turning to silver.
The spawning season range from September to February, varying with latitude and rivers. Northern populations spawn earlier than southern populations as an adaptation to slower egg development in colder water (Klemetsen et al., 2003). During spawning, females ovulate all their eggs simultaneously, then in a time windows of a few days they dig one or multiple nests where they deposit and cover the eggs after fertilization (de Gaudemar and Beall, 1998). Anadromous males do not participate in the nest construction, they rather invest their energy in aggressive competition for
18
access to fertile females (Fleming, 1996; Fleming et al., 1997, 1996) and courting behavior (de Gaudemar et al., 2000; de Gaudemar and Beall, 1999).
A different breeding strategy is applied from some male parr that achieve sexual maturation in freshwater, prior migrating to sea. They are known as “precocious parr”
or “sneaky spawners”. As the name suggests, those fish adopt an elusive breeding strategy given their smaller size compared to adult smolt. Avoiding physical competition with bigger smolt males and courtship with females, they dart into the gravel nest at oviposition to ejaculate. It has been estimated that precocious parr can contribute up to 40% of total egg fertilization (Fleming, 1996). While maturation in freshwater occur in rare cases in female Atlantic salmon (Klemetsen et al., 2003), early maturation may occur from 10 to nearly 100 % of the male population (Baum et al., 2004; Heinimaa and Erkinaro, 2004; Myers et al., 1986). Precocious parr can mature again in the following season or undergo smoltification and migrate to sea to continue their life cycle (Fleming, 1996).
Salmon males compete for fertilization of the eggs. Given the external fecundation, sperm competition is a common phenomenon among salmon, where the ejaculates of two or more males compete to fertilize a female’s ova, especially in the case of sneaky spawners (Mjølnerød et al., 1998; Stockley et al., 1997). Mature parr invest more in gonads and sperm quality, compared to anadromous males, showing higher gonadosomatic index (GSI 5-10% versus 2-6%; Fleming, 1998), spermatozoa concentration and mobility and longer spermatozoa lifespan (Daye and Glebe, 1984;
Gage et al., 1995; Vladić and Järvi, 2001). Differently from Pacific salmon, the Atlantic salmon is iteroparous, meaning that it is not genetically programmed to die after spawning. Some individuals may return to the ocean after the breeding season and repeat this cycle several times during their life span. However, due to the high mortality rate caused from predators, exhaustion and diseases, most fish survive to spawning only once or twice (Fleming, 1996).
While salmon farming is a prosperous industry that produced approximately 1.180.000 tonnes of fish in 2016, wild salmon population are facing a constant decline since 1980s, due to a variety of factors, including habitat destruction, genetic pollution, sea lice infection, acid rainfalls and environmental changes. Despite the efforts aimed to improve the quality of wild populations, the number of salmons that has returned annually from the sea to the coast of Norway has more than halved from 1983-1986 to
19
2012-2015 (55% reduction) with an estimated number of 470.000 wild individuals in 2016 (Thorstad and Forseth, 2017). As recently reported from a quality assessment on 148 Norwegian wild stocks (Thorstad et al., 2017), only 29 (20%) were classified as good quality stocks. Among the remaining, 42 (28%) were classified as moderate, 14 (9%) as bad, and 63 (43%) as very bad, according to the quality standard for wild salmon adopted under the Nature Diversity act in 2013.
The brain-pituitary-gonad (BPG) axis
Puberty is the process of morphological, physiological, and behavioural changes through which an individual becomes capable for the first time of sexual reproduction.
In male teleosts, it is marked by the onset of spermatogenesis (Schulz et al., 2010; Schulz and Miura, 2002). The regulation of sexual maturation is under control of the brain- pituitary-gonad (BPG) axis (Christensen et al., 2012; Weltzien et al., 2004; Zohar et al., 2010) (Fig. 3). Activation of the BPG axis is characterized by increased release of gonadotropin-releasing hormone (Gnrh), which stimulates synthesis and release of the two gonadotropins, follicle-stimulating hormone (Fsh) and luteinizing hormone (Lh) from the pituitary gland. These two hormones activate steroidogenesis and gametogenesis in the gonads (Schulz et al., 2010). This event is influenced by genetic factors and energy status (Bromage et al., 2001; Thorpe, 1989) as well as environmental cues like temperature and photoperiod (Bromage et al., 2001; Taranger et al., 2010).
The various environmental and physiological messages converge in the BPG axes under the form of neuropeptides and hormones. Those signals include leptin, kisspeptin, melatonin, gamma-aminobutyric acid (GABA), neuropeptide Y (NPY), acting directly on Gnrh release or modulating the Gnrh-induced gonadotropin production (Chang et al., 2009; Nakane and Oka, 2010; Navarro and Tena-Sempere, 2012; Oakley et al., 2009; Trudeau et al., 2000; Yaron et al., 2003; Zohar et al., 2010). The stimulatory action of Gnrh is countered by the inhibitory effects of dopamine in several teleost species (Dufour et al., 2010). Additionally, gonadotropin inhibitory hormone (GnIH) has been proven to decrease gonadotropin release in mammals and birds (Tsutsui, 2009), although stimulatory effects has been demonstrated in teleosts (Biran et al., 2014a). The gonadotropin production may also be influenced by a number of neurotransmitters and hormones directly at the pituitary level trough positive and negative feedback mechanisms.
20
Fig. 3 Brain-pituitary-gonadal (BPG) axis. The integrated signals originating from environmental and endogenous factors may activate the production and release of gonadotropin-releasing hormone (Gnrh). The stimulatory effects of Gnrh are exerted on the pituitary via specific Gnrh receptors leading to the secretion of the gonadotropins, follicle-stimulating hormone (Fsh) and luteinizing hormone (Lh), while dopamine can inhibit the stimulatory effects of Gnrh. Fsh and Lh, released into the blood stream, trigger the synthesis of steroid hormones (steroidogenesis) and the production of gametes (gametogenesis) after the binding to their specific receptors located in the gonads. Sex steroids affect all levels of the BPG axis through positive and negative feedback mechanisms and stimulate or inhibit the production of gonadotropins.
Once released into the blood stream, the gonadotropins may bind to specific receptors in the gonads (Fshr and Lhr) (Maugars and Schmitz, 2008, 2006; Schulz et al., 2010) activating the production of sex steroids (steroidogenesis) and mature gametes (gametogenesis). The sex steroids, including testosterone (T), 11-keto-testosterone (11KT), 17β-estradiol (E2), 17α,20β-dihydroxy-4-pregnen-3-one (DHP) can exert positive or negative feedback on all levels of the BPG axis depending on species, sex, reproductive status, and period of the year (Nagahama, 1994).
Pituitary morphology
The pituitary gland (Fig. 4), or hypophysis, is an endocrine organ responsible for the production of hormones involved in the control of different physiological and behavioural processes, such as growth, reproduction, migration, and maintenance of
21
homeostasis (Weltzien et al., 2004). It is divided in two main compartments: the adenohypophysis (or anterior pituitary), originating from the Rathke’s pouch, an ectodermal up-growth from the anterior roof of the embryonic oral cavity, and the neurohypophysis (or posterior pituitary), originating from a down-growth of the floor of the diencephalon (Wingstrand, 1966). According to the nomenclature proposed from Green, (1951), the teleost adenohypophysis can be divided into the anteriorly located pars distalis (PD) and the posteriorly located pars intermedia (PI). The PD can be further parted into the rostral pars distalis (RPD) and the proximal pars distalis (PPD). In addition to hormone producing cells, the adenohypophysis includes also non-hormonal cells such as folliculostellate cells (Fauquier et al., 2001). The neurohypophysis comprises the pars nervosa (PN) and is composed of nerve terminal from hypothalamic neuroendocrine cells, but also supportive cells named pituicytes (Ferrandino and Grimaldi, 2008). In mammals, the PN is located posterior to the adenohypophysis, while in teleosts it is often dorsally located to the adenohypophysis (Norris and Carr, 2013).
Differently from mammals, where a hypothalamo-hypophysial portal system is responsible for the transport of neurohormonal messages, in teleosts neurons directly connect the pituitary through the hypophysial stalk and the PN (Ball, 1981; Pogoda and Hammerschmidt, 2007). Whereas hormone-producing cells in the adult tetrapod pituitary are disposed in a mosaic pattern (Doerr-Schott, 1976; Voss and Rosenfeld, 1992), in teleost the different cell types are located in distinct pituitary portions, preserving the embryogenic compartmental organization (Pogoda and Hammerschmidt, 2007; Schreibman et al., 1973). The PPD hosts the gonadotropes (producing Fsh- or Lh), somatotropes (producing growth hormone, Gh) and thyrotropes (producing thyroid stimulating hormone, Tsh), while the RPD hosts the corticotropes (producing adrenocorticotropic hormone, Acth) and lactotropes (prolactin producing cells, Prl). The PI is the location of melanotropes (producing melanocyte stimulating hormone, α-Msh) and somatolactotropes (producing somatolactin, Sl) (Levavi-Sivan et al., 2010; Weltzien et al., 2004; Zohar et al., 2010).
22
Fig. 4 Schematic representation of pituitary hormone localization in a teleost (Nile tilapia). The rostral pars distalis (RPD) shows regions producing adrenocorticotropic hormone (Acth; red) and prolactin (Prl; yellow).
The proximal pars distalis (PPD) produces growth hormone (Gh; dark blue), thyroid-stimulating hormone β- subunit (Tshb; light blue), luteinizing hormone β-subunit (Lhb; light green) and follicle-stimulating hormone β- subunit (Fshb; dark green). The pars intermedia (PI) produces melanocyte-stimulating hormone (α-Msh; pink) and somatolactin (Sl; dark violet). The pars nervosa (PN) is represented in grey (reprinted with permission from Kasper et al., (2006)).
Early light microscopy studies (Olivereau, 1976) showed in PPD of salmonids the presence of two distinct gonadotrope cell types; conversely to tetrapods, where both hormones are produced from the same cells (Childs et al., 1986; Liu et al., 1988).
Later studies confirmed the presence of two distinct cell types producing Lh and Fsh in salmonids, via immunohistochemistry and in situ hybridization (Naito et al., 1993, 1991; Nozaki et al., 1990). This organization was confirmed in numerous teleosts including Bluefin tuna (Thunnus thynnus; Kagawa et al., 1998), Nile tilapia (Oreochromis niloticus; Parhar et al., 2002) Atlantic halibut (Hippoglossus hippoglossus; Weltzien et al., 2003) and Medaka (Oryzias latipes; Kanda et al., 2011). A minor portion of gonadotrope cells producing both Lh and Fsh have been reported in teleost (Kasper et al., 2006;
Pandolfi et al., 2006; Pilar García Hernández et al., 2002), as well as gonadotrope cells producing only one hormone in mammals (Pope et al., 2006; Schulz et al., 2006) suggesting a certain level of plasticity in gonadotrope cells.
GnRH
The decapeptide gonadotropin-releasing hormone (GnRH) was first characterized in mammals in 1971 (mGnRH; Amoss et al., 1971; Matsuo et al., 1971) with the primary structure identified as pGlu-His-Trp-Ser-Tyr-Gly-Leu-Arg-Pro-Gly-NH2.
23
Approximately ten years later, the first non-mammalian GnRH was identified in chicken (cGnRH-I and cGnRH-II; Miyamoto et al., 1982; Miyamoto and Hasegawa, 1984). In 1983, the first teleost Gnrh was isolated in Chum salmon, (Oncorhynchus keta; sGnrh;
Sherwood et al., 1983). Many forms of Gnrh were, since then, characterized in teleosts, including medaka (mdGnrh; Okubo et al., 2000), seabream (Sparus aurata; sbGnrh Powell et al., 1994); African catfish (Clarias gariepinus; cfGnrh Bogerd et al., 1992);
Pacific herring (Clupea harengus pallasi; hrGnrh; Carolsfeld et al., 2000) and spiny dogfish (Squalus acanthias; dfGnrh; Lovejoy et al., 1992). All the GnRH described to date, are decapeptides with perfectly conserved residues (1-Glu, 4-Ser, 9-Pro, and 10-Gly), well conserved residues (2-His and 3-Trp) and residues variable between the different forms (For review see Weltzien et al., 2004).
In vertebrates, the genes encoding GnRH share a common structure composed of four exons and three introns. The second, third and a section of the fourth exon encode for the GnRH prepro-hormone, consisting of a signal peptide (21-23 amino acids), a cleavage site (Gly-Lys-Arg), the GnRH decapeptide, and a GnRH associated peptide (GAP, 40-60 amino acids). The 5’- and3’-UTR are encoded in exons 1 and 4 respectively (Aleström et al., 1992; Chow et al., 1998; Fernald and White, 1999). The signal peptide allows the transport of the protein to the Golgi apparatus (after cleavage by a signal peptidase), where the peptide is converted in the mature form and concentrated in secretory granules. Together with the GAP, is transported to the axon terminal for its release (Andersen et al., 1988; Rangaraju et al., 1991). The GAP of the different GnRH is the most divergent region both within and among species (Roch et al., 2011).
The different forms of GnRH were initially named accordingly to the species they were first isolated, despite the fact that they can be present in other species and, as later discovered, multiple forms can be present in the same species. A new classification of the GnRH variants was proposed on the basis of phylogenetic analysis and localization of the site of expression (Fernald and White, 1999; White et al., 1995). Phylogenetic analysis clustered the different GnRH form in three groups named GnRH1, GnRH2 and GnRH3. The GnRH1 lineage includes mGnRH, cGnRH-I, and several forms isolated in fish, such as mdGnrh, hrGnrh, sbGnrh, cfGnrh and dfGnrh. The second lineage include the form found in all vertebrates, the cGnRH-II, while the GnRH3 group includes the sGnrh form. The three groups also show characteristic localization in the brain. In
24
teleost fish, the localization of Gnrh neurons was clarified in European seabass (Dicentrarchus labrax), where Gnrh1 and Gnrh3 neurons are located in olfactory bulb, ventral telencephalon and preoptic area, and Gnrh2 neurons are restricted to the midbrain tegmentum (González-Martínez et al., 2002, 2001). This organization was later confirmed in other species including lake whitefish (Coregonus clupeaformis;
Vickers et al., 2004), Cichlasoma dimerus (Pandolfi et al., 2005), Atlantic croaker (Micropogonias undulates; Mohamed et al., 2005) and medaka (Okubo et al., 2006).
The expression of two different forms of GnRH was detected in the brain of mammals, birds, reptiles (Millar, 2003), eels (Dufour et al., 1993), and more ancient teleosts including salmonids (Okuzawa et al., 1990), while three different forms are expressed in modern teleosts (Gothilf, 1996; Powell et al., 1994) and amphibians (Yoo et al., 2000). Exceptions have been detected in an ancient teleost evolving before the salmonids, the Pacific herring (Carolsfeld et al., 2000) and in a basal salmonid, the lake whitefish (Adams et al., 2002; Vickers et al., 2004), expressing three distinct Gnrh.
Those findings suggest the loss of one form in later evolving salmonids.
In teleosts producing three forms of Gnrh, the first lineage, Gnrh1, is considered to be the main responsible for the production of gonadotropins. Gnrh1 neurons were proven to directly innervate the pituitary gland, ending in the PPD or in proximity to it (where the gonadotrope cells are located) in seabream (Gothilf, 1996), Nile tilapia (Parhar et al., 1998) and European seabass (González-Martínez et al., 2004b). In the pituitary gland of maturing perciform fish, Gnrh1 (sbGnrh) is the most abundant form detected via immunoassays (Gothilf et al., 1997; Holland et al., 1998; Powell et al., 1994).
Furthermore, in Pacific herring and seabream, the ability of Gnrh1 to stimulate release of gonadotropins has been shown both in vivo and in vitro (Carolsfeld et al., 2000; Zohar et al., 1995).
As mentioned before, two forms of Gnrh are generally found in salmonids. Gnrh2 (cGnrh-II) and Gnrh3 (sGnrh) were identified in the brain of masu salmon (Oncorhynchus masou, Amano et al., 1991), sockeye salmon (Oncorhynchus nerka, Kitani et al., 2003), and rainbow trout (Oncorhynchus mykiss, Okuzawa et al., 1990). sGnrh neurons are located in the ventral forebrain from the olfactory nerve to the preoptic area innervating directly the pituitary, while cGnrh-II were found only in the midbrain tegmentum showing no innervation in the pituitary (Amano et al., 1997). Increases in gnrh mRNA in the forebrain coincide with increased gonadosomatic index and are
25
associated with precocious maturation in males (Amano et al., 1997; Ando et al., 2001) and during final gonadal maturation (Onuma et al., 2005). Furthermore, sGnrh was proven to induce Fsh secretion in vitro in coho salmon (Oncorhynchus kisutch, Dickey and Swanson, 2000). Gnrh3 (sGnrh) is therefore considered the main activator of gonadotropin secretion in salmonids.
GnRH receptors
GnRH acts via specific receptors (GnRHR) belonging to the rhodopsin β sub- family of G-protein coupled membrane receptors (Lethimonier et al., 2004; Strader et al., 1995). The protein structure is composed of seven α-helical transmembrane domains, connected by three intracellular and three extracellular loops where highly conserved amino acids form the ligand binding pocket, interaction sites with G-proteins and glycosylation sites (Millar et al., 2012).
The complexity of the GnRH system is reflected by the multiple forms on GnRHR characterized in vertebrates (Zohar et al., 2010). Despite the numerous studies conducted on GnRHR to date, the nomenclature of the different forms is still missing a clearly defined classification. The proposed systems, divided the GnRHR in two (Flanagan et al., 2007; Lethimonier et al., 2004; Moncaut, 2005), three (Levavi-Sivan et al., 2005; Millar et al., 2004) or four (Ikemoto et al., 2004; Ikemoto and Park, 2005; Kim et al., 2011) types, each composed of several subgroups. All the proposed schemes however, split teleost Gnrhr in two major groups. According to the classification proposed from Hildahl et al., 2011, (Fig. 5) the GnRHR are divided in two types. The Type I forms two subgroups, including mammalian (Type IA) and non-mammalian receptors (Type IB); and the Type II, also forming two subgroups, with fish and frog receptors on one side (Type IIB) and all the other tetrapods on the other side (Type IIA).
Mammalian Type IA receptors lack the C-terminal cytoplasmatic tail. The absence of the intracellular tail confer slower desensitization and internalization to the receptors (Hislop et al., 2001); Type II receptors maintain the C-terminal cytoplasmatic tail conferring faster desensitization and internalization to the receptor (McArdle et al., 2002).
26
Figure 5 Phylogenetic organization of vertebrate GnRHRs. Proposed from Hildahl and colleagues. The receptors are clustered in four main groups, Type IA, Type IB, Type IIA and Type IIB. Species are identified as mammalian (mam), non-mammalian tetrapod (nmt) and tetrapod (tet) clads. Teleost fish are identified by superorder: Acanthopterygii (ac); Elopomorpha (el); Ostariophysi (os); Paracanthopterygii (pa);
Protacanthopterigii (pr). (Printed with permission from Hildahl et al., 2011)
Differently from mammals that have one or two GnRHR (Hapgood et al., 2005), up to five different Gnrhr have been cloned in some teleost fish, such as European seabass (Moncaut, 2005) and spotted green pufferfish (Tetraodon nigroviridis, Ikemoto and Park, 2005). Many Gnrhr has been cloned in teleosts species including medaka (Okubo et al., 2001), European eel (Peñaranda et al., 2013), chub mackerel (Scomber
27
japonicus, Lumayno et al., 2017). In teleost pituitary, different Gnrhr may be expressed in the same cell type (Parhar et al., 2005; Strandabø et al., 2013), and the same Gnrhr may be expressed in different cell types (von Krogh et al., 2017). Nonetheless, Gnrhr may also be preferentially expressed in one cell type. For instance, gnrhr2a mRNA was identified in all Lh-producing cells and just a small number of Fsh-producing cells in European seabass (González-Martínez et al., 2004a).
Intracellular signaling pathways activated by Gnrh
Gnrhr in teleosts stimulate fshb and lhb expression through distinct intracellular pathways activated via G-proteins (for review see Chang and Pemberton, 2017; Naor, 2009; Yaron et al., 2003). Although the majority of the studies were conducted in mammals, the teleost Gnrh signal transduction pathways have been studied extensively in goldfish (Carassius auratus; Chang et al., 2009, 2000) and Nile tilapia (Gur et al., 2002;
Levavi-Sivan and Yaron, 1989).
The coupling with Gq/11 following ligand binding on the Gnrhr, activates phospholipase C (PLC) that breaks phosphatidylinositol (3,4,5)-trisphosphate (PIP3) in inositol triphosphate (IP3) and diacylglycerol (DAG). The increase of these two second messengers together with cytosolic Ca2+, activates protein kinase C (PKC). In parallel, activation of Gs stimulates adenylate cyclase (AC) activity, leading to cyclic adenosine monophosphate (cAMP) formation and activation of protein kinase A (PKA). PKA and PKC converge on MAPK/ERK (mitogen-activated protein kinases/extracellular signal- related kinases) pathways, ultimately binding to the promotor region of gpa and lhb activating their expression. The coupling with Gs protein stimulates AC activity leading to the production of cAMP and activation of PKA. PKA can either trigger fshb expression via cAMP responsive element (CRE) promoter on fshb gene or phosphorylate other MAPK cascades, including jun N-terminal kinases (JNK) that, in turn, bind AP-1 site on the promoter gene (Yaron et al., 2003) (Fig. 6)
28
Figure 6 Gnrh signalling pathways in a teleost (tilapia). (a) Proposed pathway for coordinated regulation of tilapia gpα and lhb following Gnrhr activation. Coupling with Gq stimulates phospholipase C (PLC) to produce inositol trisphosphate (IP3) and diacylglycerol (DAG). Increase of these two second messengers together with cytosolic Ca2+ activates protein kinase C (PKC). In parallel, coupling with Gs stimulates adenylate cyclase (AC) activity, increasing intracellular cAMP levels and activating protein kinase A (PKA). Both PKC and PKA converge at the level of (RAF) phosphorylating MAPK kinase (MEK) and sub sequentially the extracellular signal- regulated kinase (ERK). MAPK enters the nucleus acting on transcription factors (X) inducing gpa and lhb expression. (b) Proposed pathway for tilapia fshb regulation. Coupling with Gs stimulating AC activity and cAMP formation. PKA phosphorylate cAMP-response-element-binding protein (CREB) which activate the cAMP responsive element (CRE) on fshb promoter or act via MEKK/JNK pathway stimulating the AP1 fshb promoter.
Furthermore PKA may induce the MEK/ERK pathway acting on CRE or an unknown (X) promoter activating fshb expression.
The increase of cytosolic Ca2+ activates cascades of intracellular events as well as hormone release by exocytosis from the cells (Zhang et al., 2011). The increase in
29
intracellular Ca2+ is dependent on both extracellular influx and release from endoplasmatic reticulum (ER), regulated trough L-type voltage sensitive Ca2+ channels (VSCC; Chang et al., 2000; Hodne et al., 2013; Levavi-Sivan and Yaron, 1989) and Ca2+
activated K+ channels (KCa; Stojilkovic and Catt, 1995). In medaka, the role of VSCC is only partially responsible for the total Ca2+ influx, while the presence of many KCa
channels was detected in lhb cells, especially on of the big-conductance (BK) KCa channel (Strandabø et al., 2013). In Atlantic cod (Gadus morhua), Lh cells mainly express small- conductance (SK) KCa channels, while Fsh cells express mainly BK- KCa channels (Hodne et al., 2013), indicating both differences among species and within Lh and Fsh producing cell in the same species.
Gonadotropins
The gonadotropins are heterodimeric glycoprotein hormones belonging to the cysteine-knot growth factor superfamily. LH and FSH share a common α-subunit together with thyroid stimulating hormone (TSH) and human chorionic gonadotropin (HCG), non-covalently linked to a specific β-subunit conferring the biological activity (Pierce and Parsons, 1981; Swanson et al., 2003), where each subunit is encoded by a different gene (Fiddes and Talmadge, 1984). The first gonadotropin, LH, was purified in mammals in the late 1950s (Squire and Li, 1959), while FSH was purified almost a decade later in ovine and human pituitaries (Papkoff et al., 1967; Roos, 1968). In teleosts, all the functions ascribed to the gonadotropins, including steroidogenesis, gametogenesis, final oocyte maturation and spermiation, were initially associated to a single gonadotropin, GtH (for review see Burzawa-Gerard and Physiskogig, 1982). The presence of two distinct gonadotropins in teleosts, initially named GtH-I and GtH-II, was confirmed in chum salmon (Suzuki et al., 1988d, 1988b, 1988c) and coho salmon (Swanson et al., 1991). According to the similarity with mammalian counterparts, a resolution was adopted at the Sixth International Symposium on the Reproductive Physiology of fish, Bergen 1999 to use the term Fsh for GtH-I and Lh for GtH-II.
Gonadotropin plasma levels during sexual maturation were extensively studied in male salmonids. Fsh, already detectable in the blood in immature fish, increase significantly during early stages of maturation at the onset of spermatogenesis and again at later stages during spermatogonial proliferation and spermiation. Lh, on the other hand, is very low or undetectable during initial stages of maturation and increases
30
in later stages, during spermiation (Campbell et al., 2003; Gomez et al., 1999; Planas and Swanson, 1995; Prat et al., 1996; Suzuki et al., 1988a; Swanson et al., 1991).
Gonadotropins exert their biological activity through specific receptors. In mammals Lh regulates Leydig cell sex steroid production, while Fsh regulates Sertoli cell activities, including paracrine support of germ cell development (Huhtaniemi and Themmen, 2005). On the other hand, gonadotropin biological activity in teleosts appear less clearly distinct. Contrary to mammals, where FSHR and LHR are highly specific to their cognate hormones, in teleost this specificity is less apparent. Functional studies were conducted on gonadotropin receptors from several species including African catfish (García-López et al., 2009); Japanese eel (Anguilla japonica; Kazeto et al., 2008), coho salmon (Miwa et al., 1994; Yan et al., 1992) and Atlantic salmon (Andersson et al., 2009). A common feature of Fshr is to be activated from Fsh but also to Lh when exposed at high, but still physiological, concentrations. Conversely, Lhr were proved to be highly specific to Lh showing no cross-activation with Fsh at physiological concentrations.
In salmonids, Fsh and Lh stimulate the production of T and 11KT with comparable efficiency (Planas et al., 1993), while Lh is a more potent stimulator of DHP production during final maturation and spawning (Planas and Swanson, 1995). In African catfish and Japanese eel the presence of Fshr has been detected in both Leydig and Sertoli cells, while Lhr was detected only in Leydig cells (García-López et al., 2009;
Ohta et al., 2007). In coho salmon Fshr were detected in Sertoli cells but the authors did not exclude a possible localization also in Leydig cells, while Lhr were only detected in Leydig cells (Miwa et al., 1994). Taken together, receptor localization and pharmacological data suggest that Leydig cell steroidogenic activity is regulated by both Lh and Fsh, while the functions of Sertoli cells are mainly regulated from Fsh, although during the spawning season, high concentrations of Lh may activate Fshr.
Melatonin
All living beings have adapted their physiological and behavioural functions to daily and annual fluctuations of environmental cues. Photoperiod, the alternation of light (L) and darkness (D) during the 24 hours, is the main and most reliable of these cues. It is considered as a “noise free” signal, since its variations are consistent, year after year, in the lifespan of an animal. Temperature, water salinity, light spectrum and
31
food availability are some among other features that can shape biological rhythms.
Despite their periodicity, those cues are more subjected to short time variation and therefore considered “noisy” signals (Cowan et al., 2017; Falcón et al., 2010).
A circadian system (Fig. 7) includes all the components by which the photic signal is perceived and converted into a timed hormonal and nervous output (Falcón et al., 2007). In mammals, the photic information is conveyed from the eyes to the suprachiasmatic nuclei of the hypothalamus (SCN), considered as the master clock site, generating the pulse for all circadian rhythms. From here the information is transported to the pineal gland, a vesicle attached by a stalk to the roof of the diencephalon, responsible for the production of circulating melatonin (Simonneaux, 2003). Melatonin (5-methoxy-N-acetyltryptamine) is the main output of the vertebrate circadian clocks, enabling the synchronization of environmental conditions (using photoperiod as the main driver) with physiological, metabolic and behavioural processes, including reproduction (Falcón et al., 2011). In teleosts, the circadian system is organized in a less linear way, as a network of more interconnected circadian units, with retina and pineal having always a central position (Falcón et al., 2007).
Figure 7 Photoperiodic and circadian regulation of neuroendocrine functions. (A) Schematic representation of the linear flow leading to the rhythmic production of melatonin in mammals. Non-visual information collected from the eyes, travel via the retinohypotalamic tract to the suprachiasmatic nucleus (SCN) of the hypothalamus (blue arrow). The photoperiodic signals influence the synchronization of the circadian activity of the SCN clocks, which in turn influence the cyclic melatonin production from the pineal gland through a multisynaptic pathway (blue arrows). Melatonin feeds back to the SCN and modulates seasonal
32
neuroendocrine physiological and behavioural functions acting on the pars tuberalis of the pituitary gland, brain regions and peripheral tissues. (B) Photoneuroendocrine regulation in fish. Photoreceptor cells in the pineal and retina perceive light stimulus (yellow arrows) thus synchronizing their internal clocks. Light may also act on other possible photosensitive and circadian structures in the preoptic area (POA) of the ventral diencephalon. The retina and the pineal produce both a neural (blue arrow) and a hormonal (red arrow) response to the photoperiodic information. The neural response reaches the ventral diencephalon via the retinohypotalamic and the pineal tracts. This response provides information regarding day length and variation in ambient illumination. The hormonal response is carried out from melatonin (red arrows) whose production reflects day length and season. Retinal melatonin acts as an autocrine/paracrine factor, while pineal melatonin is released in the blood and cerebrospinal fluid acting through melatonin receptors on specific targets (red circles). Melatonin receptors have been identified on locations, such as the preoptic area (POA), involved in the control of pituitary functions and in the pituitary gland itself (adapted from Falcón et al., 2010 with permission).
Melatonin is an indoleamine hormone produced during the night phase from tryptophan, which is successively transformed into serotonin (5-hydroxytryptamine) from a two-step enzymatic reaction. The conversion from serotonin to melatonin is controlled by two enzymes. The first, arylalkylamine N-acetyltransferase (AANAT), is responsible for the formation of N-acetylserotonin via acetylation of serotonin. The second, hydroxyindole-O-methyltransferase (HIOMT), produce melatonin via O- methylation of N-acetylserotonin (Falcón et al., 2011, 2006; Klein et al., 1997). Unlike other vertebrates, two aanat genes are expressed in teleosts, aanat1 and aanat2, probably as a result of genome duplication (Falcón et al., 2007). The former is mainly expressed in retina, brain and peripheral tissues and the latter is mainly produced in the pineal (Cazaméa-Catalan et al., 2014, 2012; Falcón et al., 2010; Paulin et al., 2015).
Aanat2 is the rate-limiting enzyme in the nocturnal production of melatonin by photoreceptor cells in the pineal gland (Falcón, 1999; Falcón et al., 2011; Ziv et al., 2005), since both Aanat2 mRNA expression and protein activity are inhibited by light (Falcón et al., 2010, 2001; Migaud et al., 2006). While the timing of aanat2 expression and protein functionality is controlled by photoperiod, the amplitude of Aanat2 activity and therefore melatonin production, is influenced by temperature in combination with other external and internal factors (Falcón et al., 2010). In a number of teleosts species, the maximum Aanat2 activity coincide with the species-specific optimal water temperature interval as a result of evolutionary adaptation (Falcón et al., 2009). Other external factors such as spectral quality of light, salinity, fish species and also fish
33
populations within the same species may have a strong impact on melatonin production (Bayarri et al., 2002; Brüning et al., 2016; Lopez-Patino et al., 2011; Migaud et al., 2006;
Oliveira et al., 2007; Porter et al., 2001; Skulstad et al., 2013; Vera et al., 2005).
Furthermore, internal factors such as steroids (Chattoraj et al., 2009; Lopez-Patino et al., 2014) and melatonin itself, can influence melatonin production (Falcón et al., 2011).
The release of melatonin in the plasma during the night phase is a conserved characteristic common to all vertebrates, regardless of a diurnal or nocturnal nature of the animal. Three different types (A, B and C) of plasma melatonin profiles can be distinguished in teleosts, as in vertebrates in general. Type-A profiles are characterized by a peak of melatonin during late dark phase and are typical of gadoid species such as Atlantic cod and haddock (Melanogrammus aeglefinus) (Davie et al., 2007; Porter et al., 2000). Plasma melatonin in Type-B profiles peak during the middle of the dark phase, as in Nile tilapia (Martinez-Chavez et al., 2008). Type-C profiles immediately peak at the onset of the dark period, maintaining high levels until morning light. The latter profile is characteristic of salmonid species, and a range of other teleosts (Acuña-Castroviejo et al., 2014; Ceinos et al., 2005; Migaud et al., 2010).
The circadian clock system is a molecular feedback loop where two heterodimers, PER/CRY (repressor) and BMAL/CLOCK (activator) drive the rhythmic expression of a number of genes (including aanat2) in phase with solar time, allowing the anticipation of environmental changes (Appelbaum and Gothilf, 2006; Reppert and Weaver, 2002; Zilberman-Peled et al., 2007). In tropical teleosts, adapted to a constant photoperiod throughout the year, the phase of the rhythm is stably locked to 12L/12D cycle, while in temperate teleosts the phase is adjusted daily (Ziv et al., 2005). Many teleost species exhibit circadian rhythms of melatonin release in constant darkness (DD), including zebrafish, pike (Esox lucius), white sucker (Catostomus commersonii), and ayu (Plecoglossus altivelis) (Bolliet et al., 1996; Cahill, 1996; Iigo et al., 2004; Kazimi and Cahill, 1999; Zachmann et al., 1992b, 1992a). Salmonids on the other hand, are lacking this system and show a constant melatonin production under DD (Falcón et al., 2007; Gern and Greenhouse, 1988; Iigo et al., 2007; McStay et al., 2014; Thibault et al., 1993).
Extrapineal melatonin production has been detected in several peripheral tissues such as retina and gut. Interestingly, teleost retinal melatonin production is not limited to the dark phase, as seen in zebrafish and goldfish (Cahill, 1996; Iigo et al.,
34
1997a) but can also occur during the day (Besseau et al., 2006; Gern et al., 1978; Iigo et al., 1997b). The melatonin produced from the retina is not released into the blood stream, but act as a local autocrine/paracrine signal (Besseau et al., 2006; Falcón et al., 2010; Ping et al., 2008; Sauzet et al., 2008). Nonetheless, light perceptions from the eyes may influence circulating melatonin levels in an indirect way, through stimulation of pineal melatonin release in some teleost (Martinez-Chavez et al., 2008; Martinez- Chavez and Migaud, 2009; Migaud et al., 2007; Nikaido et al., 2009). Melatonin production has been detected in the gastrointestinal tract (GIT) of several teleost species (Isorna et al., 2017) including rainbow trout (Muñoz-Pérez et al., 2016) goldfish, carp and European seabass, likely regulated by feeding time and tryptophan availability (J. Y. Choi et al., 2016; Y. J. Choi et al., 2016; Herrero et al., 2007; Mukherjee and Maitra, 2015). The role of melatonin production in the gut may be related to the regulation of the digestive process (Vera et al., 2007).
Melatonin receptors
Melatonin acts via specific melatonin receptors (Mtnr) belonging to the G- protein coupled receptor superfamily (Brydon et al., 1999). In vertebrates, the receptors are divided in three sub-groups: Mtnr1A (Mel1a or MT1), Mtnr1B (Mel1b or MT2) and Mtnr1C (Mel1c or GPR50). An additional Mtnr1A was reported in some teleost species (Ikegami et al., 2009; Mazurais et al., 1999; Reppert et al., 1995), however the origin of the two teleost Mtnr1A paralogues was not identified.
Mtnr act through different intracellular pathways: Mtnr1A and Mtnr1B activate the cyclic adenosine monophosphate/protein kinase A (cAMP/PKA) pathway coupling with Gi-protein. This results in inhibition of adenylyl cyclase (AC) and reduced cAMP formation (Rimler et al., 2006). Mtnr1A and Mtnr1C activate the phospholipase C/protein kinase C (PLC/PKC) pathway via Gq-proteins (Balík et al., 2004), while Mtnr1B activate the cyclic guanosine monophosphate (cGMP) pathway (Huang et al., 2005). Mtnr1C has lost the capacity to respond to melatonin in therian mammals (Dubocovich et al., 2010; Gautier et al., 2018).
The involvement of melatonin in a number of different physiological and behavioural functions is reflected from the wide distribution of Mtnr in vertebrate nervous and peripheral tissues (Witt-Enderby et al., 2003). The presence of Mtnr in the pituitary gland is of particular interest for the purpose of this thesis. Mtnr has been
35
identified in the pars tuberalis (PT) of the mammalian pituitary (Ebisawa et al., 1994;
Schuster et al., 2001). In teleosts the presence of either mtnr or melatonin binding sites was detected in the pituitary of several species including goldfish (Ikegami et al., 2009), European seabass (Sauzet et al., 2008), Senegalese sole (Solea senegalensis; Confente et al., 2010), and salmonids like chum salmon, pike, rainbow trout (Falcón et al., 2003;
Gaildrat and Falcón, 2002; Shi et al., 2004), but not Atlantic salmon (Ekström and Vanĕcek, 1992). Whether melatonin effects on reproduction in teleosts result from direct effects on gonadotrope cells, indirect effects through other components of the BPG axes, or a combination of both is still not clear, however in some species melatonin was proven to influence gonadotropin levels in vivo (Khan and Thomas, 1996; Sébert et al., 2008).
Testes morphology
In teleosts, the testicles are elongated structures covered by a thin connective capsule, the tunica albuginea, located in the body cavity between two lobes divided by a sectum. They share the general structure common to all vertebrates, composed of two main compartments, the tubular and the intertubular compartment (Schulz et al., 2010). The intertubular compartment is composed of connective tissue, blood vessels, fibroblasts and Leydig cells. The Leydig cells are involved in the production of steroids, influencing both spermatogenesis and secondary sexual characteristics (Koulish et al., 2002). Teleosts produce two main androgens: 11KT, acting as a direct activator of spermatogenesis (Borg, 1994; Cavaco et al., 1998; Miura et al., 1991a) and T, influencing reproduction through positive and negative feedback in several tissues, including hypothalamus (Amano et al., 1994; Goos et al., 1986) and pituitary (Dufour et al., 1983;
Montero et al., 1995; Xiong et al., 1993). The tubular compartment is composed of a basement membrane enclosing Sertoli cells and germ cells. Sertoli cells provide physical support but also paracrine factors needed for germ cell proliferation and differentiation (DiNapoli and Capel, 2008). Sertoli cells are involved also in phagocytosis of apoptotic germ cells and residual sperm after the spawning season (Almeida et al., 2008; Vilela et al., 2003). During the onset of meiotic division, Sertoli cells form a blood-testis barrier around the dividing germinal cells (Batlouni et al., 2009).
In amniote vertebrates (reptiles, birds, mammals) Sertoli cells arrest their proliferation at puberty and support waves of dividing germ cells at different
36
developmental stages. In anamniote vertebrates (fish and amphibians) the tubular compartment is organized in spermatocysts where a genetically determined number of Sertoli cells (Matta et al., 2002) surround a clone of germ cells all at the same developmental stage (Billard et al., 1982; Engel and Callard, 2007; Pudney, 1995).
Vertebrate testis can be classified upon the morphology of the germinal compartment and the distribution of the germ cells. In teleosts, two types of spermatogonial distribution have been described. In the first type (restricted), spermatogonia are located in the testis periphery near the tunica albuginea. The cysts migrate towards the spermatic ducts at the centre of the testis during the meiotic cell division. This organization is typical of higher teleosts (Parenti and Grier, 2004). In the second type (unrestricted), spermatogonia are distributed all along the length of the testis and cysts do not migrate during meiotic divisions. This organization is typical of lower teleosts including salmonids (Parenti and Grier, 2004).
Spermatogenesis
Spermatogenesis is a highly coordinated process where diploid spermatogonia differentiate to produce mature haploid spermatozoa. The general process is conserved in vertebrates and can be divided in three main phases: the mitotic or spermatogonial phase, the meiotic phase, and the spermiogenic phase (for review see Schulz et al., 2010)
During the first phase, mitotic proliferation of spermatogonial stem cells lead to the production of new stem cells and differentiated spermatogonia. Following the nomenclature proposed from Schulz et al. (2010, Fig 8), undifferentiated type A spermatogonia (Aund) transform into differentiated type A spermatogonia (Adiff) followed by type B spermatogonia, with a fixed number of generations for each species (Ando et al., 2000). Primary spermatocytes are formed after the final mitotic division.
Simultaneously Sertoli cells undergo a defined number of mitotic divisions (Nagahama, 2000). The germ cell proliferation can be induced in teleosts from 11KT (Miura et al., 1991b; Nader et al., 1999) and insulin like growth factor I (IGF-I; Loir and Le Gac, 1994;
Nader et al., 1999), while Sertoli cell proliferation is stimulated by FSH in both mammals (Kumar et al., 1999) and teleosts (Lejeune et al., 1996).
